Upload folder using huggingface_hub
Browse files
slots/8/checkpoint-5758/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TwinyForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 3072,
|
| 11 |
+
"max_position_embeddings": 128,
|
| 12 |
+
"model_type": "twiny",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 1,
|
| 16 |
+
"num_key_value_heads": 3,
|
| 17 |
+
"qk_norm": true,
|
| 18 |
+
"rezero_init": 1.0,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.0.0",
|
| 22 |
+
"use_cache": false,
|
| 23 |
+
"vocab_size": 32000
|
| 24 |
+
}
|
slots/8/checkpoint-5758/trainer_state.json
ADDED
|
@@ -0,0 +1,2050 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.4464415162974498,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5758,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 2.617016627215141e-05,
|
| 14 |
+
"grad_norm": 122.33340454101562,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 419.9996032714844,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0005234033254430282,
|
| 21 |
+
"grad_norm": 90.82410430908203,
|
| 22 |
+
"learning_rate": 4.999999511980259e-05,
|
| 23 |
+
"loss": 428.18197471217104,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0010468066508860564,
|
| 28 |
+
"grad_norm": 59.52641296386719,
|
| 29 |
+
"learning_rate": 4.999997943828395e-05,
|
| 30 |
+
"loss": 366.3947265625,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0015702099763290846,
|
| 35 |
+
"grad_norm": 41.61277389526367,
|
| 36 |
+
"learning_rate": 4.999995294193233e-05,
|
| 37 |
+
"loss": 325.6545166015625,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0020936133017721128,
|
| 42 |
+
"grad_norm": 46.67556381225586,
|
| 43 |
+
"learning_rate": 4.999991563075919e-05,
|
| 44 |
+
"loss": 310.3648193359375,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.002617016627215141,
|
| 49 |
+
"grad_norm": 47.65824890136719,
|
| 50 |
+
"learning_rate": 4.999986750478065e-05,
|
| 51 |
+
"loss": 303.49443359375,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.003140419952658169,
|
| 56 |
+
"grad_norm": 62.748138427734375,
|
| 57 |
+
"learning_rate": 4.999980856401755e-05,
|
| 58 |
+
"loss": 322.389990234375,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.007327646556202395,
|
| 63 |
+
"grad_norm": 48.20169448852539,
|
| 64 |
+
"learning_rate": 4.999895523943923e-05,
|
| 65 |
+
"loss": 288.82497336647725,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.008374453207088451,
|
| 70 |
+
"grad_norm": 68.66603088378906,
|
| 71 |
+
"learning_rate": 4.99986329623214e-05,
|
| 72 |
+
"loss": 293.87509765625,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.009421259857974507,
|
| 77 |
+
"grad_norm": 61.69401931762695,
|
| 78 |
+
"learning_rate": 4.999826742820622e-05,
|
| 79 |
+
"loss": 324.10361328125,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.010468066508860565,
|
| 84 |
+
"grad_norm": 56.778385162353516,
|
| 85 |
+
"learning_rate": 4.999785863772618e-05,
|
| 86 |
+
"loss": 274.875537109375,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.01151487315974662,
|
| 91 |
+
"grad_norm": 71.3816146850586,
|
| 92 |
+
"learning_rate": 4.999740659158865e-05,
|
| 93 |
+
"loss": 281.357080078125,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.012561679810632677,
|
| 98 |
+
"grad_norm": 75.08697509765625,
|
| 99 |
+
"learning_rate": 4.9996911290575844e-05,
|
| 100 |
+
"loss": 263.852783203125,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.013608486461518732,
|
| 105 |
+
"grad_norm": 53.91341781616211,
|
| 106 |
+
"learning_rate": 4.999637273554481e-05,
|
| 107 |
+
"loss": 258.546630859375,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.01465529311240479,
|
| 112 |
+
"grad_norm": 56.39162063598633,
|
| 113 |
+
"learning_rate": 4.9995790927427446e-05,
|
| 114 |
+
"loss": 270.5470703125,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.015702099763290844,
|
| 119 |
+
"grad_norm": 62.93777084350586,
|
| 120 |
+
"learning_rate": 4.99951658672305e-05,
|
| 121 |
+
"loss": 252.8383056640625,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.016748906414176902,
|
| 126 |
+
"grad_norm": 61.51578903198242,
|
| 127 |
+
"learning_rate": 4.9994497556035567e-05,
|
| 128 |
+
"loss": 287.8167724609375,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.01779571306506296,
|
| 133 |
+
"grad_norm": 62.75235366821289,
|
| 134 |
+
"learning_rate": 4.9993785994999074e-05,
|
| 135 |
+
"loss": 258.673779296875,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.018842519715949014,
|
| 140 |
+
"grad_norm": 56.2258186340332,
|
| 141 |
+
"learning_rate": 4.999303118535229e-05,
|
| 142 |
+
"loss": 252.9609375,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.01988932636683507,
|
| 147 |
+
"grad_norm": 54.998130798339844,
|
| 148 |
+
"learning_rate": 4.9992233128401314e-05,
|
| 149 |
+
"loss": 241.9411376953125,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.02093613301772113,
|
| 154 |
+
"grad_norm": 63.37522888183594,
|
| 155 |
+
"learning_rate": 4.999139182552709e-05,
|
| 156 |
+
"loss": 249.7280517578125,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.021982939668607183,
|
| 161 |
+
"grad_norm": 60.007144927978516,
|
| 162 |
+
"learning_rate": 4.9990507278185374e-05,
|
| 163 |
+
"loss": 236.941015625,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.02302974631949324,
|
| 168 |
+
"grad_norm": 66.06304168701172,
|
| 169 |
+
"learning_rate": 4.998957948790677e-05,
|
| 170 |
+
"loss": 240.55878363715277,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0240765529703793,
|
| 175 |
+
"grad_norm": 68.9736099243164,
|
| 176 |
+
"learning_rate": 4.998860845629671e-05,
|
| 177 |
+
"loss": 247.988720703125,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.025123359621265353,
|
| 182 |
+
"grad_norm": 61.50578308105469,
|
| 183 |
+
"learning_rate": 4.998759418503544e-05,
|
| 184 |
+
"loss": 233.025537109375,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.02617016627215141,
|
| 189 |
+
"grad_norm": 54.96660614013672,
|
| 190 |
+
"learning_rate": 4.998653667587801e-05,
|
| 191 |
+
"loss": 243.004736328125,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.06804234327362181,
|
| 196 |
+
"grad_norm": 50.733734130859375,
|
| 197 |
+
"learning_rate": 4.990903999473118e-05,
|
| 198 |
+
"loss": 250.31749378551137,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.07065935647645342,
|
| 203 |
+
"grad_norm": 48.06721496582031,
|
| 204 |
+
"learning_rate": 4.9901899191179905e-05,
|
| 205 |
+
"loss": 243.69541015625,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.07327636967928504,
|
| 210 |
+
"grad_norm": 54.863311767578125,
|
| 211 |
+
"learning_rate": 4.989448913413268e-05,
|
| 212 |
+
"loss": 225.0319580078125,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07589338288211664,
|
| 217 |
+
"grad_norm": 60.36458969116211,
|
| 218 |
+
"learning_rate": 4.988680990371125e-05,
|
| 219 |
+
"loss": 222.7537353515625,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.07851039608494825,
|
| 224 |
+
"grad_norm": 57.877986907958984,
|
| 225 |
+
"learning_rate": 4.987886158294783e-05,
|
| 226 |
+
"loss": 241.103515625,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.08112740928777985,
|
| 231 |
+
"grad_norm": 64.6783676147461,
|
| 232 |
+
"learning_rate": 4.987064425778417e-05,
|
| 233 |
+
"loss": 215.971142578125,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.08374442249061147,
|
| 238 |
+
"grad_norm": 50.828765869140625,
|
| 239 |
+
"learning_rate": 4.986215801707066e-05,
|
| 240 |
+
"loss": 213.7265380859375,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.08636143569344307,
|
| 245 |
+
"grad_norm": 55.68048858642578,
|
| 246 |
+
"learning_rate": 4.9853402952565383e-05,
|
| 247 |
+
"loss": 216.94306640625,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08897844889627468,
|
| 252 |
+
"grad_norm": 54.125267028808594,
|
| 253 |
+
"learning_rate": 4.984437915893305e-05,
|
| 254 |
+
"loss": 204.139990234375,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0915954620991063,
|
| 259 |
+
"grad_norm": 57.27487564086914,
|
| 260 |
+
"learning_rate": 4.9835086733744074e-05,
|
| 261 |
+
"loss": 202.9946533203125,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.0942124753019379,
|
| 266 |
+
"grad_norm": 65.67598724365234,
|
| 267 |
+
"learning_rate": 4.9825525777473424e-05,
|
| 268 |
+
"loss": 207.21923828125,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.09682948850476951,
|
| 273 |
+
"grad_norm": 71.1091537475586,
|
| 274 |
+
"learning_rate": 4.98156963934996e-05,
|
| 275 |
+
"loss": 204.4235107421875,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.09944650170760111,
|
| 280 |
+
"grad_norm": 60.16761779785156,
|
| 281 |
+
"learning_rate": 4.9805598688103496e-05,
|
| 282 |
+
"loss": 213.841748046875,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.19596194862803082,
|
| 287 |
+
"grad_norm": Infinity,
|
| 288 |
+
"learning_rate": 4.99999986487106e-05,
|
| 289 |
+
"loss": 453.1419677734375,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2009866139774675,
|
| 294 |
+
"grad_norm": 119.09779357910156,
|
| 295 |
+
"learning_rate": 4.9999459486180746e-05,
|
| 296 |
+
"loss": 453.07587890625,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.20601127932690422,
|
| 301 |
+
"grad_norm": 102.9168701171875,
|
| 302 |
+
"learning_rate": 4.9997837968095394e-05,
|
| 303 |
+
"loss": 465.0671875,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2110359446763409,
|
| 308 |
+
"grad_norm": 113.77227020263672,
|
| 309 |
+
"learning_rate": 4.999513551586019e-05,
|
| 310 |
+
"loss": 412.81953125,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2160606100257776,
|
| 315 |
+
"grad_norm": 101.17367553710938,
|
| 316 |
+
"learning_rate": 4.999135224633215e-05,
|
| 317 |
+
"loss": 416.068896484375,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.22108527537521427,
|
| 322 |
+
"grad_norm": 105.45433044433594,
|
| 323 |
+
"learning_rate": 4.9986757181733314e-05,
|
| 324 |
+
"loss": 453.432958984375,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.22610994072465096,
|
| 329 |
+
"grad_norm": 108.6885986328125,
|
| 330 |
+
"learning_rate": 4.998086683157904e-05,
|
| 331 |
+
"loss": 408.4590576171875,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.23113460607408765,
|
| 336 |
+
"grad_norm": 103.64579772949219,
|
| 337 |
+
"learning_rate": 4.997389628112563e-05,
|
| 338 |
+
"loss": 379.013232421875,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.23615927142352433,
|
| 343 |
+
"grad_norm": 88.69979858398438,
|
| 344 |
+
"learning_rate": 4.996584583178737e-05,
|
| 345 |
+
"loss": 388.706591796875,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.24118393677296102,
|
| 350 |
+
"grad_norm": 91.30870056152344,
|
| 351 |
+
"learning_rate": 4.995671583167459e-05,
|
| 352 |
+
"loss": 397.874267578125,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2462086021223977,
|
| 357 |
+
"grad_norm": 107.91041564941406,
|
| 358 |
+
"learning_rate": 4.9946506675578595e-05,
|
| 359 |
+
"loss": 384.0211669921875,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.2512332674718344,
|
| 364 |
+
"grad_norm": 89.37213134765625,
|
| 365 |
+
"learning_rate": 4.993521880495458e-05,
|
| 366 |
+
"loss": 380.9727783203125,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.2562579328212711,
|
| 371 |
+
"grad_norm": 110.07548522949219,
|
| 372 |
+
"learning_rate": 4.992285270790253e-05,
|
| 373 |
+
"loss": 352.642529296875,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.26128259817070776,
|
| 378 |
+
"grad_norm": 95.92185974121094,
|
| 379 |
+
"learning_rate": 4.990940891914619e-05,
|
| 380 |
+
"loss": 345.220703125,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.26630726352014444,
|
| 385 |
+
"grad_norm": 94.39682006835938,
|
| 386 |
+
"learning_rate": 4.989488802000982e-05,
|
| 387 |
+
"loss": 344.5824462890625,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.27133192886958113,
|
| 392 |
+
"grad_norm": 101.4179916381836,
|
| 393 |
+
"learning_rate": 4.987929063839317e-05,
|
| 394 |
+
"loss": 333.548291015625,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.2763565942190178,
|
| 399 |
+
"grad_norm": 113.27001190185547,
|
| 400 |
+
"learning_rate": 4.9862617448744245e-05,
|
| 401 |
+
"loss": 315.02744140625,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2813812595684545,
|
| 406 |
+
"grad_norm": 82.49633026123047,
|
| 407 |
+
"learning_rate": 4.984486917203022e-05,
|
| 408 |
+
"loss": 312.227978515625,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2864059249178912,
|
| 413 |
+
"grad_norm": 95.27268981933594,
|
| 414 |
+
"learning_rate": 4.982604657570619e-05,
|
| 415 |
+
"loss": 348.5288330078125,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.29143059026732787,
|
| 420 |
+
"grad_norm": 102.54169464111328,
|
| 421 |
+
"learning_rate": 4.980615047368202e-05,
|
| 422 |
+
"loss": 325.8141357421875,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.2964552556167646,
|
| 427 |
+
"grad_norm": 91.8856430053711,
|
| 428 |
+
"learning_rate": 4.9785181726287174e-05,
|
| 429 |
+
"loss": 312.114892578125,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3014799209662013,
|
| 434 |
+
"grad_norm": 84.40894317626953,
|
| 435 |
+
"learning_rate": 4.976314124023346e-05,
|
| 436 |
+
"loss": 291.1855224609375,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.306504586315638,
|
| 441 |
+
"grad_norm": 79.73738098144531,
|
| 442 |
+
"learning_rate": 4.9740029968575874e-05,
|
| 443 |
+
"loss": 290.4890380859375,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.31152925166507467,
|
| 448 |
+
"grad_norm": 84.9390869140625,
|
| 449 |
+
"learning_rate": 4.971584891067134e-05,
|
| 450 |
+
"loss": 279.4406005859375,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.31655391701451135,
|
| 455 |
+
"grad_norm": 103.37007141113281,
|
| 456 |
+
"learning_rate": 4.969059911213554e-05,
|
| 457 |
+
"loss": 278.745849609375,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.32157858236394804,
|
| 462 |
+
"grad_norm": 85.26111602783203,
|
| 463 |
+
"learning_rate": 4.966428166479768e-05,
|
| 464 |
+
"loss": 286.081298828125,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3266032477133847,
|
| 469 |
+
"grad_norm": 85.8335952758789,
|
| 470 |
+
"learning_rate": 4.963689770665327e-05,
|
| 471 |
+
"loss": 259.804248046875,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3316279130628214,
|
| 476 |
+
"grad_norm": 85.86480712890625,
|
| 477 |
+
"learning_rate": 4.960844842181494e-05,
|
| 478 |
+
"loss": 246.61513671875,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3366525784122581,
|
| 483 |
+
"grad_norm": 73.1225814819336,
|
| 484 |
+
"learning_rate": 4.957893504046123e-05,
|
| 485 |
+
"loss": 283.9971923828125,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.3416772437616948,
|
| 490 |
+
"grad_norm": 70.62370300292969,
|
| 491 |
+
"learning_rate": 4.9548358838783354e-05,
|
| 492 |
+
"loss": 237.363037109375,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.34670190911113147,
|
| 497 |
+
"grad_norm": 76.33558654785156,
|
| 498 |
+
"learning_rate": 4.951672113893009e-05,
|
| 499 |
+
"loss": 267.32021484375,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.35172657446056815,
|
| 504 |
+
"grad_norm": 87.06673431396484,
|
| 505 |
+
"learning_rate": 4.948402330895056e-05,
|
| 506 |
+
"loss": 255.372705078125,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.35675123981000484,
|
| 511 |
+
"grad_norm": 75.33609771728516,
|
| 512 |
+
"learning_rate": 4.9450266762735066e-05,
|
| 513 |
+
"loss": 223.300732421875,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.3617759051594415,
|
| 518 |
+
"grad_norm": 80.52996063232422,
|
| 519 |
+
"learning_rate": 4.9415452959954e-05,
|
| 520 |
+
"loss": 240.35537109375,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3668005705088782,
|
| 525 |
+
"grad_norm": 72.82714080810547,
|
| 526 |
+
"learning_rate": 4.937958340599467e-05,
|
| 527 |
+
"loss": 216.26923828125,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.3718252358583149,
|
| 532 |
+
"grad_norm": 64.65206909179688,
|
| 533 |
+
"learning_rate": 4.934265965189626e-05,
|
| 534 |
+
"loss": 211.017578125,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3768499012077516,
|
| 539 |
+
"grad_norm": 83.19200897216797,
|
| 540 |
+
"learning_rate": 4.930468329428271e-05,
|
| 541 |
+
"loss": 244.0173583984375,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.38187456655718827,
|
| 546 |
+
"grad_norm": 91.22970581054688,
|
| 547 |
+
"learning_rate": 4.9265655975293693e-05,
|
| 548 |
+
"loss": 206.4826416015625,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.38689923190662495,
|
| 553 |
+
"grad_norm": 70.5400161743164,
|
| 554 |
+
"learning_rate": 4.9225579382513656e-05,
|
| 555 |
+
"loss": 227.105810546875,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.39192389725606164,
|
| 560 |
+
"grad_norm": 83.25762176513672,
|
| 561 |
+
"learning_rate": 4.918445524889876e-05,
|
| 562 |
+
"loss": 232.9232177734375,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.3969485626054983,
|
| 567 |
+
"grad_norm": 69.92899322509766,
|
| 568 |
+
"learning_rate": 4.914228535270201e-05,
|
| 569 |
+
"loss": 201.66182861328124,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.401973227954935,
|
| 574 |
+
"grad_norm": 61.95820999145508,
|
| 575 |
+
"learning_rate": 4.909907151739633e-05,
|
| 576 |
+
"loss": 190.1841064453125,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4069978933043717,
|
| 581 |
+
"grad_norm": 92.61400604248047,
|
| 582 |
+
"learning_rate": 4.905481561159575e-05,
|
| 583 |
+
"loss": 205.5430908203125,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.41202255865380844,
|
| 588 |
+
"grad_norm": 82.73475646972656,
|
| 589 |
+
"learning_rate": 4.900951954897456e-05,
|
| 590 |
+
"loss": 194.57945556640624,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.4170472240032451,
|
| 595 |
+
"grad_norm": 71.5935287475586,
|
| 596 |
+
"learning_rate": 4.8963185288184574e-05,
|
| 597 |
+
"loss": 183.76383056640626,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.4220718893526818,
|
| 602 |
+
"grad_norm": 71.0921401977539,
|
| 603 |
+
"learning_rate": 4.891581483277046e-05,
|
| 604 |
+
"loss": 186.3758056640625,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4270965547021185,
|
| 609 |
+
"grad_norm": 59.311614990234375,
|
| 610 |
+
"learning_rate": 4.886741023108307e-05,
|
| 611 |
+
"loss": 211.0236083984375,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.4321212200515552,
|
| 616 |
+
"grad_norm": 69.83146667480469,
|
| 617 |
+
"learning_rate": 4.8817973576190915e-05,
|
| 618 |
+
"loss": 188.28985595703125,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.43714588540099186,
|
| 623 |
+
"grad_norm": 62.22911834716797,
|
| 624 |
+
"learning_rate": 4.876750700578959e-05,
|
| 625 |
+
"loss": 173.9123291015625,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.44217055075042855,
|
| 630 |
+
"grad_norm": 59.1146240234375,
|
| 631 |
+
"learning_rate": 4.871601270210939e-05,
|
| 632 |
+
"loss": 188.10584716796876,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.44719521609986523,
|
| 637 |
+
"grad_norm": 79.95441436767578,
|
| 638 |
+
"learning_rate": 4.866349289182095e-05,
|
| 639 |
+
"loss": 174.2633544921875,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4522198814493019,
|
| 644 |
+
"grad_norm": 68.70292663574219,
|
| 645 |
+
"learning_rate": 4.860994984593892e-05,
|
| 646 |
+
"loss": 167.2957275390625,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.4572445467987386,
|
| 651 |
+
"grad_norm": 62.22568130493164,
|
| 652 |
+
"learning_rate": 4.855538587972379e-05,
|
| 653 |
+
"loss": 160.5239990234375,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4622692121481753,
|
| 658 |
+
"grad_norm": 63.13420104980469,
|
| 659 |
+
"learning_rate": 4.8499803352581795e-05,
|
| 660 |
+
"loss": 166.31607666015626,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.467293877497612,
|
| 665 |
+
"grad_norm": 75.40135192871094,
|
| 666 |
+
"learning_rate": 4.844320466796286e-05,
|
| 667 |
+
"loss": 159.2119873046875,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.47231854284704866,
|
| 672 |
+
"grad_norm": 57.91975021362305,
|
| 673 |
+
"learning_rate": 4.8385592273256675e-05,
|
| 674 |
+
"loss": 167.14599609375,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.47734320819648535,
|
| 679 |
+
"grad_norm": 72.13136291503906,
|
| 680 |
+
"learning_rate": 4.832696865968686e-05,
|
| 681 |
+
"loss": 167.04654541015626,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.48236787354592203,
|
| 686 |
+
"grad_norm": 70.31126403808594,
|
| 687 |
+
"learning_rate": 4.826733636220332e-05,
|
| 688 |
+
"loss": 151.86346435546875,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.4873925388953587,
|
| 693 |
+
"grad_norm": 63.09219741821289,
|
| 694 |
+
"learning_rate": 4.820669795937248e-05,
|
| 695 |
+
"loss": 137.3025146484375,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.4924172042447954,
|
| 700 |
+
"grad_norm": 72.1025619506836,
|
| 701 |
+
"learning_rate": 4.8145056073265945e-05,
|
| 702 |
+
"loss": 144.692138671875,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.4974418695942321,
|
| 707 |
+
"grad_norm": 63.844993591308594,
|
| 708 |
+
"learning_rate": 4.8082413369347003e-05,
|
| 709 |
+
"loss": 145.94105224609376,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5024665349436688,
|
| 714 |
+
"grad_norm": 65.97188568115234,
|
| 715 |
+
"learning_rate": 4.8018772556355436e-05,
|
| 716 |
+
"loss": 154.27625732421876,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5074912002931055,
|
| 721 |
+
"grad_norm": 65.91217041015625,
|
| 722 |
+
"learning_rate": 4.795413638619035e-05,
|
| 723 |
+
"loss": 137.7441162109375,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5125158656425421,
|
| 728 |
+
"grad_norm": 61.9046745300293,
|
| 729 |
+
"learning_rate": 4.788850765379119e-05,
|
| 730 |
+
"loss": 132.897412109375,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5175405309919788,
|
| 735 |
+
"grad_norm": 56.1972770690918,
|
| 736 |
+
"learning_rate": 4.782188919701692e-05,
|
| 737 |
+
"loss": 133.23360595703124,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5225651963414155,
|
| 742 |
+
"grad_norm": 64.27227783203125,
|
| 743 |
+
"learning_rate": 4.775428389652326e-05,
|
| 744 |
+
"loss": 120.74188232421875,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5275898616908522,
|
| 749 |
+
"grad_norm": 71.39759826660156,
|
| 750 |
+
"learning_rate": 4.768569467563812e-05,
|
| 751 |
+
"loss": 125.49366455078125,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5326145270402889,
|
| 756 |
+
"grad_norm": 57.68914794921875,
|
| 757 |
+
"learning_rate": 4.761612450023526e-05,
|
| 758 |
+
"loss": 148.16036376953124,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5376391923897256,
|
| 763 |
+
"grad_norm": 49.9764289855957,
|
| 764 |
+
"learning_rate": 4.754557637860597e-05,
|
| 765 |
+
"loss": 120.12298583984375,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5426638577391623,
|
| 770 |
+
"grad_norm": 64.86238861083984,
|
| 771 |
+
"learning_rate": 4.747405336132903e-05,
|
| 772 |
+
"loss": 118.1151123046875,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5476885230885989,
|
| 777 |
+
"grad_norm": 89.1286849975586,
|
| 778 |
+
"learning_rate": 4.740155854113877e-05,
|
| 779 |
+
"loss": 126.01204833984374,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5527131884380356,
|
| 784 |
+
"grad_norm": 68.27723693847656,
|
| 785 |
+
"learning_rate": 4.732809505279136e-05,
|
| 786 |
+
"loss": 118.21903076171876,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5577378537874723,
|
| 791 |
+
"grad_norm": 56.56875228881836,
|
| 792 |
+
"learning_rate": 4.725366607292926e-05,
|
| 793 |
+
"loss": 110.47099609375,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.562762519136909,
|
| 798 |
+
"grad_norm": 62.396175384521484,
|
| 799 |
+
"learning_rate": 4.717827481994384e-05,
|
| 800 |
+
"loss": 106.74912109375,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5677871844863457,
|
| 805 |
+
"grad_norm": 77.86981964111328,
|
| 806 |
+
"learning_rate": 4.710192455383623e-05,
|
| 807 |
+
"loss": 110.87667236328124,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5728118498357824,
|
| 812 |
+
"grad_norm": 53.221561431884766,
|
| 813 |
+
"learning_rate": 4.702461857607634e-05,
|
| 814 |
+
"loss": 119.735498046875,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5778365151852191,
|
| 819 |
+
"grad_norm": 54.47859573364258,
|
| 820 |
+
"learning_rate": 4.694636022946012e-05,
|
| 821 |
+
"loss": 108.375927734375,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5828611805346557,
|
| 826 |
+
"grad_norm": 56.99690628051758,
|
| 827 |
+
"learning_rate": 4.6867152897964984e-05,
|
| 828 |
+
"loss": 103.08570556640625,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.5878858458840925,
|
| 833 |
+
"grad_norm": 55.26579666137695,
|
| 834 |
+
"learning_rate": 4.679103005487893e-05,
|
| 835 |
+
"loss": 101.62680053710938,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5929105112335292,
|
| 840 |
+
"grad_norm": 60.823219299316406,
|
| 841 |
+
"learning_rate": 4.67099820913045e-05,
|
| 842 |
+
"loss": 98.35703125,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.5979351765829659,
|
| 847 |
+
"grad_norm": 51.73524856567383,
|
| 848 |
+
"learning_rate": 4.6627995364103185e-05,
|
| 849 |
+
"loss": 94.52880859375,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6029598419324026,
|
| 854 |
+
"grad_norm": 59.78525161743164,
|
| 855 |
+
"learning_rate": 4.65450734184717e-05,
|
| 856 |
+
"loss": 98.74429321289062,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6079845072818393,
|
| 861 |
+
"grad_norm": 52.353214263916016,
|
| 862 |
+
"learning_rate": 4.6461219840046654e-05,
|
| 863 |
+
"loss": 97.32471313476563,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.613009172631276,
|
| 868 |
+
"grad_norm": 46.18357849121094,
|
| 869 |
+
"learning_rate": 4.6376438254749485e-05,
|
| 870 |
+
"loss": 92.0089111328125,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6180338379807127,
|
| 875 |
+
"grad_norm": 64.59222412109375,
|
| 876 |
+
"learning_rate": 4.629073232862966e-05,
|
| 877 |
+
"loss": 101.53438720703124,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.6230585033301493,
|
| 882 |
+
"grad_norm": 65.1756820678711,
|
| 883 |
+
"learning_rate": 4.620410576770618e-05,
|
| 884 |
+
"loss": 95.05399780273437,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.628083168679586,
|
| 889 |
+
"grad_norm": 56.11637496948242,
|
| 890 |
+
"learning_rate": 4.611656231780731e-05,
|
| 891 |
+
"loss": 100.00510864257812,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.6331078340290227,
|
| 896 |
+
"grad_norm": 48.66570281982422,
|
| 897 |
+
"learning_rate": 4.6028105764408604e-05,
|
| 898 |
+
"loss": 98.039013671875,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.6381324993784594,
|
| 903 |
+
"grad_norm": 47.48982238769531,
|
| 904 |
+
"learning_rate": 4.593873993246923e-05,
|
| 905 |
+
"loss": 81.84201049804688,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.6431571647278961,
|
| 910 |
+
"grad_norm": 41.5953369140625,
|
| 911 |
+
"learning_rate": 4.5848468686266554e-05,
|
| 912 |
+
"loss": 91.85083618164063,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6481818300773328,
|
| 917 |
+
"grad_norm": 73.18049621582031,
|
| 918 |
+
"learning_rate": 4.575729592922906e-05,
|
| 919 |
+
"loss": 88.3189208984375,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.6532064954267695,
|
| 924 |
+
"grad_norm": 50.01223373413086,
|
| 925 |
+
"learning_rate": 4.566522560376756e-05,
|
| 926 |
+
"loss": 82.32785034179688,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.6582311607762061,
|
| 931 |
+
"grad_norm": 42.86427688598633,
|
| 932 |
+
"learning_rate": 4.557226169110471e-05,
|
| 933 |
+
"loss": 85.24454956054687,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.6632558261256428,
|
| 938 |
+
"grad_norm": 61.30537414550781,
|
| 939 |
+
"learning_rate": 4.547840821110287e-05,
|
| 940 |
+
"loss": 76.753076171875,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6682804914750795,
|
| 945 |
+
"grad_norm": 48.51723861694336,
|
| 946 |
+
"learning_rate": 4.5383669222090276e-05,
|
| 947 |
+
"loss": 83.67793579101563,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.6733051568245162,
|
| 952 |
+
"grad_norm": 58.5217170715332,
|
| 953 |
+
"learning_rate": 4.528804882068555e-05,
|
| 954 |
+
"loss": 74.97109985351562,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.6783298221739529,
|
| 959 |
+
"grad_norm": 46.08441925048828,
|
| 960 |
+
"learning_rate": 4.519155114162056e-05,
|
| 961 |
+
"loss": 84.21328125,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.6833544875233896,
|
| 966 |
+
"grad_norm": 50.42264938354492,
|
| 967 |
+
"learning_rate": 4.509418035756163e-05,
|
| 968 |
+
"loss": 71.0347412109375,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6883791528728262,
|
| 973 |
+
"grad_norm": 46.333213806152344,
|
| 974 |
+
"learning_rate": 4.499594067892911e-05,
|
| 975 |
+
"loss": 69.8539794921875,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.6934038182222629,
|
| 980 |
+
"grad_norm": 51.47473907470703,
|
| 981 |
+
"learning_rate": 4.489683635371531e-05,
|
| 982 |
+
"loss": 70.3105224609375,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.6984284835716996,
|
| 987 |
+
"grad_norm": 50.31460952758789,
|
| 988 |
+
"learning_rate": 4.4796871667300835e-05,
|
| 989 |
+
"loss": 75.20213012695312,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.7034531489211363,
|
| 994 |
+
"grad_norm": 43.928733825683594,
|
| 995 |
+
"learning_rate": 4.4696050942269216e-05,
|
| 996 |
+
"loss": 69.22677612304688,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.708477814270573,
|
| 1001 |
+
"grad_norm": 56.49867248535156,
|
| 1002 |
+
"learning_rate": 4.459437853822007e-05,
|
| 1003 |
+
"loss": 69.10577392578125,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.7135024796200097,
|
| 1008 |
+
"grad_norm": 50.505577087402344,
|
| 1009 |
+
"learning_rate": 4.449185885158056e-05,
|
| 1010 |
+
"loss": 64.09615478515624,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.7185271449694464,
|
| 1015 |
+
"grad_norm": 44.39089584350586,
|
| 1016 |
+
"learning_rate": 4.4388496315415275e-05,
|
| 1017 |
+
"loss": 83.494921875,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.723551810318883,
|
| 1022 |
+
"grad_norm": 57.29201126098633,
|
| 1023 |
+
"learning_rate": 4.4284295399234545e-05,
|
| 1024 |
+
"loss": 69.22371215820313,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.7285764756683197,
|
| 1029 |
+
"grad_norm": 57.06948471069336,
|
| 1030 |
+
"learning_rate": 4.417926060880117e-05,
|
| 1031 |
+
"loss": 69.6142578125,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.7336011410177564,
|
| 1036 |
+
"grad_norm": 43.410362243652344,
|
| 1037 |
+
"learning_rate": 4.407339648593564e-05,
|
| 1038 |
+
"loss": 64.617236328125,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.7386258063671931,
|
| 1043 |
+
"grad_norm": 52.63591766357422,
|
| 1044 |
+
"learning_rate": 4.396670760831965e-05,
|
| 1045 |
+
"loss": 58.28870849609375,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.7436504717166298,
|
| 1050 |
+
"grad_norm": 45.577003479003906,
|
| 1051 |
+
"learning_rate": 4.3859198589298206e-05,
|
| 1052 |
+
"loss": 60.7156494140625,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.7486751370660665,
|
| 1057 |
+
"grad_norm": 57.865943908691406,
|
| 1058 |
+
"learning_rate": 4.375087407768016e-05,
|
| 1059 |
+
"loss": 63.53287963867187,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.7536998024155032,
|
| 1064 |
+
"grad_norm": 44.069618225097656,
|
| 1065 |
+
"learning_rate": 4.364173875753715e-05,
|
| 1066 |
+
"loss": 63.81923828125,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.7587244677649398,
|
| 1071 |
+
"grad_norm": 45.2602653503418,
|
| 1072 |
+
"learning_rate": 4.353179734800106e-05,
|
| 1073 |
+
"loss": 62.04966430664062,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.7637491331143765,
|
| 1078 |
+
"grad_norm": 50.382720947265625,
|
| 1079 |
+
"learning_rate": 4.342105460306e-05,
|
| 1080 |
+
"loss": 60.8373046875,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.7687737984638132,
|
| 1085 |
+
"grad_norm": 40.19420623779297,
|
| 1086 |
+
"learning_rate": 4.3309515311352686e-05,
|
| 1087 |
+
"loss": 62.08988647460937,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.7737984638132499,
|
| 1092 |
+
"grad_norm": 39.89312744140625,
|
| 1093 |
+
"learning_rate": 4.319718429596139e-05,
|
| 1094 |
+
"loss": 55.46314697265625,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7788231291626866,
|
| 1099 |
+
"grad_norm": 49.28807067871094,
|
| 1100 |
+
"learning_rate": 4.308406641420343e-05,
|
| 1101 |
+
"loss": 55.56544189453125,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.7838477945121233,
|
| 1106 |
+
"grad_norm": 43.144046783447266,
|
| 1107 |
+
"learning_rate": 4.2970166557421044e-05,
|
| 1108 |
+
"loss": 56.770928955078126,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.78887245986156,
|
| 1113 |
+
"grad_norm": 66.24620056152344,
|
| 1114 |
+
"learning_rate": 4.285548965076998e-05,
|
| 1115 |
+
"loss": 53.5606689453125,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.7938971252109966,
|
| 1120 |
+
"grad_norm": 46.12450408935547,
|
| 1121 |
+
"learning_rate": 4.274004065300644e-05,
|
| 1122 |
+
"loss": 51.77293090820312,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.7989217905604333,
|
| 1127 |
+
"grad_norm": 50.26192092895508,
|
| 1128 |
+
"learning_rate": 4.262382455627274e-05,
|
| 1129 |
+
"loss": 52.784716796875,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.80394645590987,
|
| 1134 |
+
"grad_norm": 41.408443450927734,
|
| 1135 |
+
"learning_rate": 4.250684638588138e-05,
|
| 1136 |
+
"loss": 51.038412475585936,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.8089711212593067,
|
| 1141 |
+
"grad_norm": 39.26192855834961,
|
| 1142 |
+
"learning_rate": 4.2389111200097765e-05,
|
| 1143 |
+
"loss": 47.814639282226565,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.8139957866087434,
|
| 1148 |
+
"grad_norm": 47.69087219238281,
|
| 1149 |
+
"learning_rate": 4.227062408992149e-05,
|
| 1150 |
+
"loss": 48.265420532226564,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.8190204519581802,
|
| 1155 |
+
"grad_norm": 44.28941345214844,
|
| 1156 |
+
"learning_rate": 4.21513901788662e-05,
|
| 1157 |
+
"loss": 46.7813232421875,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.8240451173076169,
|
| 1162 |
+
"grad_norm": 48.63425064086914,
|
| 1163 |
+
"learning_rate": 4.2031414622738024e-05,
|
| 1164 |
+
"loss": 49.20258483886719,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.8290697826570536,
|
| 1169 |
+
"grad_norm": 41.629661560058594,
|
| 1170 |
+
"learning_rate": 4.1910702609412644e-05,
|
| 1171 |
+
"loss": 48.036032104492186,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.8340944480064902,
|
| 1176 |
+
"grad_norm": 48.02088928222656,
|
| 1177 |
+
"learning_rate": 4.1789259358610966e-05,
|
| 1178 |
+
"loss": 49.056277465820315,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.8391191133559269,
|
| 1183 |
+
"grad_norm": 44.047969818115234,
|
| 1184 |
+
"learning_rate": 4.1667090121673415e-05,
|
| 1185 |
+
"loss": 46.96828002929688,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.8441437787053636,
|
| 1190 |
+
"grad_norm": 42.91846466064453,
|
| 1191 |
+
"learning_rate": 4.154420018133286e-05,
|
| 1192 |
+
"loss": 43.54451293945313,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.8491684440548003,
|
| 1197 |
+
"grad_norm": 71.76822662353516,
|
| 1198 |
+
"learning_rate": 4.14205948514862e-05,
|
| 1199 |
+
"loss": 42.600888061523435,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.854193109404237,
|
| 1204 |
+
"grad_norm": 41.15577697753906,
|
| 1205 |
+
"learning_rate": 4.129627947696451e-05,
|
| 1206 |
+
"loss": 40.6418701171875,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.8592177747536737,
|
| 1211 |
+
"grad_norm": 33.31337356567383,
|
| 1212 |
+
"learning_rate": 4.117125943330205e-05,
|
| 1213 |
+
"loss": 46.37944030761719,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.8642424401031104,
|
| 1218 |
+
"grad_norm": 45.18565368652344,
|
| 1219 |
+
"learning_rate": 4.104554012650372e-05,
|
| 1220 |
+
"loss": 45.95015258789063,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.869267105452547,
|
| 1225 |
+
"grad_norm": 43.49892044067383,
|
| 1226 |
+
"learning_rate": 4.091912699281132e-05,
|
| 1227 |
+
"loss": 39.445840454101564,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.8742917708019837,
|
| 1232 |
+
"grad_norm": 37.17277145385742,
|
| 1233 |
+
"learning_rate": 4.0792025498468515e-05,
|
| 1234 |
+
"loss": 40.20647277832031,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.8793164361514204,
|
| 1239 |
+
"grad_norm": 43.78445816040039,
|
| 1240 |
+
"learning_rate": 4.066424113948444e-05,
|
| 1241 |
+
"loss": 39.10909729003906,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.8843411015008571,
|
| 1246 |
+
"grad_norm": 37.20656204223633,
|
| 1247 |
+
"learning_rate": 4.053577944139604e-05,
|
| 1248 |
+
"loss": 42.784820556640625,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.8893657668502938,
|
| 1253 |
+
"grad_norm": 38.29347610473633,
|
| 1254 |
+
"learning_rate": 4.040664595902916e-05,
|
| 1255 |
+
"loss": 38.08096008300781,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.8943904321997305,
|
| 1260 |
+
"grad_norm": 42.234046936035156,
|
| 1261 |
+
"learning_rate": 4.027684627625834e-05,
|
| 1262 |
+
"loss": 35.874212646484374,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.8994150975491672,
|
| 1267 |
+
"grad_norm": 61.35478973388672,
|
| 1268 |
+
"learning_rate": 4.014638600576537e-05,
|
| 1269 |
+
"loss": 40.434365844726564,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.9044397628986038,
|
| 1274 |
+
"grad_norm": 43.708946228027344,
|
| 1275 |
+
"learning_rate": 4.0015270788796567e-05,
|
| 1276 |
+
"loss": 35.79759216308594,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.9094644282480405,
|
| 1281 |
+
"grad_norm": 38.74837112426758,
|
| 1282 |
+
"learning_rate": 3.988350629491887e-05,
|
| 1283 |
+
"loss": 36.532388305664064,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.9144890935974772,
|
| 1288 |
+
"grad_norm": 34.870052337646484,
|
| 1289 |
+
"learning_rate": 3.975109822177466e-05,
|
| 1290 |
+
"loss": 38.52042846679687,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.9195137589469139,
|
| 1295 |
+
"grad_norm": 36.34028625488281,
|
| 1296 |
+
"learning_rate": 3.961805229483542e-05,
|
| 1297 |
+
"loss": 39.098443603515626,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.9245384242963506,
|
| 1302 |
+
"grad_norm": 33.49582290649414,
|
| 1303 |
+
"learning_rate": 3.948437426715409e-05,
|
| 1304 |
+
"loss": 35.11366577148438,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.9295630896457873,
|
| 1309 |
+
"grad_norm": 36.892520904541016,
|
| 1310 |
+
"learning_rate": 3.9350069919116396e-05,
|
| 1311 |
+
"loss": 34.33470153808594,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.934587754995224,
|
| 1316 |
+
"grad_norm": 37.274940490722656,
|
| 1317 |
+
"learning_rate": 3.921514505819081e-05,
|
| 1318 |
+
"loss": 31.69019775390625,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.9396124203446606,
|
| 1323 |
+
"grad_norm": 35.98411560058594,
|
| 1324 |
+
"learning_rate": 3.90796055186775e-05,
|
| 1325 |
+
"loss": 37.6157470703125,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.9446370856940973,
|
| 1330 |
+
"grad_norm": 46.616085052490234,
|
| 1331 |
+
"learning_rate": 3.894345716145598e-05,
|
| 1332 |
+
"loss": 31.594985961914062,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.949661751043534,
|
| 1337 |
+
"grad_norm": 41.32733154296875,
|
| 1338 |
+
"learning_rate": 3.8806705873731754e-05,
|
| 1339 |
+
"loss": 33.326727294921874,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.9546864163929707,
|
| 1344 |
+
"grad_norm": 31.950937271118164,
|
| 1345 |
+
"learning_rate": 3.866935756878167e-05,
|
| 1346 |
+
"loss": 32.37601928710937,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.9597110817424074,
|
| 1351 |
+
"grad_norm": 32.85530471801758,
|
| 1352 |
+
"learning_rate": 3.853141818569829e-05,
|
| 1353 |
+
"loss": 32.27983703613281,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.9647357470918441,
|
| 1358 |
+
"grad_norm": 32.26963806152344,
|
| 1359 |
+
"learning_rate": 3.8392893689133026e-05,
|
| 1360 |
+
"loss": 31.69603271484375,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.9697604124412808,
|
| 1365 |
+
"grad_norm": 38.555084228515625,
|
| 1366 |
+
"learning_rate": 3.825379006903826e-05,
|
| 1367 |
+
"loss": 29.670452880859376,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.9747850777907174,
|
| 1372 |
+
"grad_norm": 37.716068267822266,
|
| 1373 |
+
"learning_rate": 3.811411334040832e-05,
|
| 1374 |
+
"loss": 30.258895874023438,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.9798097431401541,
|
| 1379 |
+
"grad_norm": 52.04689025878906,
|
| 1380 |
+
"learning_rate": 3.7973869543019345e-05,
|
| 1381 |
+
"loss": 28.086944580078125,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.9848344084895908,
|
| 1386 |
+
"grad_norm": 39.67953109741211,
|
| 1387 |
+
"learning_rate": 3.78330647411682e-05,
|
| 1388 |
+
"loss": 30.239852905273438,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.9898590738390275,
|
| 1393 |
+
"grad_norm": 30.14206886291504,
|
| 1394 |
+
"learning_rate": 3.769170502341018e-05,
|
| 1395 |
+
"loss": 26.414932250976562,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.9948837391884642,
|
| 1400 |
+
"grad_norm": 44.91764831542969,
|
| 1401 |
+
"learning_rate": 3.754979650229575e-05,
|
| 1402 |
+
"loss": 27.337918090820313,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.9999084045379009,
|
| 1407 |
+
"grad_norm": 34.629451751708984,
|
| 1408 |
+
"learning_rate": 3.740734531410626e-05,
|
| 1409 |
+
"loss": 27.581338500976564,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.0047734320819648,
|
| 1414 |
+
"grad_norm": 30.09258270263672,
|
| 1415 |
+
"learning_rate": 3.726435761858856e-05,
|
| 1416 |
+
"loss": 24.877536010742187,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.0097980974314016,
|
| 1421 |
+
"grad_norm": 39.526817321777344,
|
| 1422 |
+
"learning_rate": 3.7120839598688686e-05,
|
| 1423 |
+
"loss": 25.269071960449217,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.0148227627808382,
|
| 1428 |
+
"grad_norm": 39.116336822509766,
|
| 1429 |
+
"learning_rate": 3.697679746028449e-05,
|
| 1430 |
+
"loss": 28.1403076171875,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.019847428130275,
|
| 1435 |
+
"grad_norm": 27.28890609741211,
|
| 1436 |
+
"learning_rate": 3.683223743191729e-05,
|
| 1437 |
+
"loss": 25.208116149902345,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.0248720934797115,
|
| 1442 |
+
"grad_norm": 24.81954002380371,
|
| 1443 |
+
"learning_rate": 3.6687165764522507e-05,
|
| 1444 |
+
"loss": 23.084710693359376,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.0298967588291483,
|
| 1449 |
+
"grad_norm": 30.29689598083496,
|
| 1450 |
+
"learning_rate": 3.654158873115944e-05,
|
| 1451 |
+
"loss": 27.75635986328125,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.034921424178585,
|
| 1456 |
+
"grad_norm": 28.4095401763916,
|
| 1457 |
+
"learning_rate": 3.639551262673994e-05,
|
| 1458 |
+
"loss": 23.95079803466797,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.0399460895280217,
|
| 1463 |
+
"grad_norm": 23.15727996826172,
|
| 1464 |
+
"learning_rate": 3.624894376775626e-05,
|
| 1465 |
+
"loss": 22.17493438720703,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.0449707548774583,
|
| 1470 |
+
"grad_norm": 27.223628997802734,
|
| 1471 |
+
"learning_rate": 3.6101888492007905e-05,
|
| 1472 |
+
"loss": 24.21749267578125,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.049995420226895,
|
| 1477 |
+
"grad_norm": 27.820846557617188,
|
| 1478 |
+
"learning_rate": 3.5954353158327555e-05,
|
| 1479 |
+
"loss": 25.064503479003907,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.0550200855763316,
|
| 1484 |
+
"grad_norm": 25.031492233276367,
|
| 1485 |
+
"learning_rate": 3.580634414630618e-05,
|
| 1486 |
+
"loss": 26.935647583007814,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.0600447509257684,
|
| 1491 |
+
"grad_norm": 37.862030029296875,
|
| 1492 |
+
"learning_rate": 3.565786785601706e-05,
|
| 1493 |
+
"loss": 22.24079132080078,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.065069416275205,
|
| 1498 |
+
"grad_norm": 33.74036407470703,
|
| 1499 |
+
"learning_rate": 3.550893070773914e-05,
|
| 1500 |
+
"loss": 21.394287109375,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.0700940816246418,
|
| 1505 |
+
"grad_norm": 27.69246482849121,
|
| 1506 |
+
"learning_rate": 3.535953914167937e-05,
|
| 1507 |
+
"loss": 22.197325134277342,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.0751187469740784,
|
| 1512 |
+
"grad_norm": 29.03541374206543,
|
| 1513 |
+
"learning_rate": 3.520969961769424e-05,
|
| 1514 |
+
"loss": 20.178665161132812,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.0801434123235152,
|
| 1519 |
+
"grad_norm": 28.151607513427734,
|
| 1520 |
+
"learning_rate": 3.505941861501038e-05,
|
| 1521 |
+
"loss": 20.440557861328124,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.0851680776729518,
|
| 1526 |
+
"grad_norm": 44.759029388427734,
|
| 1527 |
+
"learning_rate": 3.4908702631944535e-05,
|
| 1528 |
+
"loss": 22.989773559570313,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.0901927430223886,
|
| 1533 |
+
"grad_norm": 30.5145206451416,
|
| 1534 |
+
"learning_rate": 3.4757558185622404e-05,
|
| 1535 |
+
"loss": 17.94971160888672,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.0952174083718251,
|
| 1540 |
+
"grad_norm": 27.25557518005371,
|
| 1541 |
+
"learning_rate": 3.460599181169696e-05,
|
| 1542 |
+
"loss": 20.101849365234376,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.100242073721262,
|
| 1547 |
+
"grad_norm": 33.63078308105469,
|
| 1548 |
+
"learning_rate": 3.4454010064065764e-05,
|
| 1549 |
+
"loss": 20.41118621826172,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.1052667390706987,
|
| 1554 |
+
"grad_norm": 24.511587142944336,
|
| 1555 |
+
"learning_rate": 3.43016195145876e-05,
|
| 1556 |
+
"loss": 18.789512634277344,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.1102914044201353,
|
| 1561 |
+
"grad_norm": 27.190340042114258,
|
| 1562 |
+
"learning_rate": 3.414882675279832e-05,
|
| 1563 |
+
"loss": 20.354037475585937,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.1153160697695719,
|
| 1568 |
+
"grad_norm": 27.395174026489258,
|
| 1569 |
+
"learning_rate": 3.399563838562585e-05,
|
| 1570 |
+
"loss": 16.81817169189453,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.1203407351190087,
|
| 1575 |
+
"grad_norm": 24.14755630493164,
|
| 1576 |
+
"learning_rate": 3.3842061037104535e-05,
|
| 1577 |
+
"loss": 16.935992431640624,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.1253654004684455,
|
| 1582 |
+
"grad_norm": 27.58745765686035,
|
| 1583 |
+
"learning_rate": 3.368810134808873e-05,
|
| 1584 |
+
"loss": 19.668711853027343,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.130390065817882,
|
| 1589 |
+
"grad_norm": 27.285415649414062,
|
| 1590 |
+
"learning_rate": 3.3533765975965583e-05,
|
| 1591 |
+
"loss": 16.50089111328125,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.1354147311673188,
|
| 1596 |
+
"grad_norm": 26.07541847229004,
|
| 1597 |
+
"learning_rate": 3.3379061594367214e-05,
|
| 1598 |
+
"loss": 17.075025939941405,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.1404393965167554,
|
| 1603 |
+
"grad_norm": 22.186742782592773,
|
| 1604 |
+
"learning_rate": 3.322399489288212e-05,
|
| 1605 |
+
"loss": 16.471742248535158,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.1454640618661922,
|
| 1610 |
+
"grad_norm": 43.99259567260742,
|
| 1611 |
+
"learning_rate": 3.306857257676588e-05,
|
| 1612 |
+
"loss": 16.091929626464843,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.1504887272156288,
|
| 1617 |
+
"grad_norm": 22.85940170288086,
|
| 1618 |
+
"learning_rate": 3.291280136665131e-05,
|
| 1619 |
+
"loss": 16.774765014648438,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.1555133925650656,
|
| 1624 |
+
"grad_norm": 30.904027938842773,
|
| 1625 |
+
"learning_rate": 3.2756687998257706e-05,
|
| 1626 |
+
"loss": 20.361752319335938,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.1605380579145022,
|
| 1631 |
+
"grad_norm": 25.8914737701416,
|
| 1632 |
+
"learning_rate": 3.260023922209974e-05,
|
| 1633 |
+
"loss": 16.825946044921874,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.165562723263939,
|
| 1638 |
+
"grad_norm": 21.448165893554688,
|
| 1639 |
+
"learning_rate": 3.244346180319543e-05,
|
| 1640 |
+
"loss": 14.290866088867187,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.1705873886133755,
|
| 1645 |
+
"grad_norm": 23.256567001342773,
|
| 1646 |
+
"learning_rate": 3.228636252077371e-05,
|
| 1647 |
+
"loss": 14.399835205078125,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.1756120539628123,
|
| 1652 |
+
"grad_norm": 31.78139305114746,
|
| 1653 |
+
"learning_rate": 3.2128948167981224e-05,
|
| 1654 |
+
"loss": 17.567897033691406,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.180636719312249,
|
| 1659 |
+
"grad_norm": 28.218175888061523,
|
| 1660 |
+
"learning_rate": 3.1971225551588606e-05,
|
| 1661 |
+
"loss": 15.964744567871094,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.1856613846616857,
|
| 1666 |
+
"grad_norm": 23.535409927368164,
|
| 1667 |
+
"learning_rate": 3.181320149169617e-05,
|
| 1668 |
+
"loss": 13.499687194824219,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.1906860500111223,
|
| 1673 |
+
"grad_norm": 23.55646514892578,
|
| 1674 |
+
"learning_rate": 3.165488282143897e-05,
|
| 1675 |
+
"loss": 13.131063842773438,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.195710715360559,
|
| 1680 |
+
"grad_norm": 37.625816345214844,
|
| 1681 |
+
"learning_rate": 3.149627638669132e-05,
|
| 1682 |
+
"loss": 13.719525146484376,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.2007353807099956,
|
| 1687 |
+
"grad_norm": 23.7028751373291,
|
| 1688 |
+
"learning_rate": 3.133738904577083e-05,
|
| 1689 |
+
"loss": 14.057861328125,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.2057600460594324,
|
| 1694 |
+
"grad_norm": 32.64527130126953,
|
| 1695 |
+
"learning_rate": 3.1178227669141744e-05,
|
| 1696 |
+
"loss": 13.545637512207032,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.210784711408869,
|
| 1701 |
+
"grad_norm": 19.4030818939209,
|
| 1702 |
+
"learning_rate": 3.101879913911797e-05,
|
| 1703 |
+
"loss": 13.282673645019532,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.2158093767583058,
|
| 1708 |
+
"grad_norm": 20.120149612426758,
|
| 1709 |
+
"learning_rate": 3.085911034956539e-05,
|
| 1710 |
+
"loss": 13.5511474609375,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.2208340421077424,
|
| 1715 |
+
"grad_norm": 19.937030792236328,
|
| 1716 |
+
"learning_rate": 3.0699168205603816e-05,
|
| 1717 |
+
"loss": 12.457398223876954,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.2258587074571792,
|
| 1722 |
+
"grad_norm": 24.85066795349121,
|
| 1723 |
+
"learning_rate": 3.053897962330838e-05,
|
| 1724 |
+
"loss": 12.30173110961914,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.2308833728066157,
|
| 1729 |
+
"grad_norm": 21.03877067565918,
|
| 1730 |
+
"learning_rate": 3.0378551529410452e-05,
|
| 1731 |
+
"loss": 12.57455062866211,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.2359080381560525,
|
| 1736 |
+
"grad_norm": 18.235923767089844,
|
| 1737 |
+
"learning_rate": 3.02178908609982e-05,
|
| 1738 |
+
"loss": 12.81219940185547,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.2409327035054891,
|
| 1743 |
+
"grad_norm": 23.471614837646484,
|
| 1744 |
+
"learning_rate": 3.0057004565216522e-05,
|
| 1745 |
+
"loss": 12.431439208984376,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.245957368854926,
|
| 1750 |
+
"grad_norm": 22.228742599487305,
|
| 1751 |
+
"learning_rate": 2.9895899598966725e-05,
|
| 1752 |
+
"loss": 11.524738311767578,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.2509820342043625,
|
| 1757 |
+
"grad_norm": 20.493696212768555,
|
| 1758 |
+
"learning_rate": 2.9734582928605648e-05,
|
| 1759 |
+
"loss": 11.868292999267577,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.2560066995537993,
|
| 1764 |
+
"grad_norm": 33.29052734375,
|
| 1765 |
+
"learning_rate": 2.957306152964447e-05,
|
| 1766 |
+
"loss": 11.813479614257812,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.2610313649032359,
|
| 1771 |
+
"grad_norm": 19.822860717773438,
|
| 1772 |
+
"learning_rate": 2.9411342386447045e-05,
|
| 1773 |
+
"loss": 10.693547821044922,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.2660560302526727,
|
| 1778 |
+
"grad_norm": 21.911001205444336,
|
| 1779 |
+
"learning_rate": 2.9249432491927915e-05,
|
| 1780 |
+
"loss": 11.636901092529296,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.2710806956021092,
|
| 1785 |
+
"grad_norm": 24.26630973815918,
|
| 1786 |
+
"learning_rate": 2.9087338847249912e-05,
|
| 1787 |
+
"loss": 12.051991271972657,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.276105360951546,
|
| 1792 |
+
"grad_norm": 27.69570541381836,
|
| 1793 |
+
"learning_rate": 2.892506846152143e-05,
|
| 1794 |
+
"loss": 10.351026153564453,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.2811300263009828,
|
| 1799 |
+
"grad_norm": 21.411516189575195,
|
| 1800 |
+
"learning_rate": 2.8762628351493365e-05,
|
| 1801 |
+
"loss": 10.994412994384765,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.2861546916504194,
|
| 1806 |
+
"grad_norm": 21.352632522583008,
|
| 1807 |
+
"learning_rate": 2.860002554125563e-05,
|
| 1808 |
+
"loss": 10.254959869384766,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.291179356999856,
|
| 1813 |
+
"grad_norm": 17.76573371887207,
|
| 1814 |
+
"learning_rate": 2.8437267061933527e-05,
|
| 1815 |
+
"loss": 10.489431762695313,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.2962040223492928,
|
| 1820 |
+
"grad_norm": 18.212268829345703,
|
| 1821 |
+
"learning_rate": 2.8274359951383628e-05,
|
| 1822 |
+
"loss": 10.117932891845703,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.3012286876987296,
|
| 1827 |
+
"grad_norm": 20.227031707763672,
|
| 1828 |
+
"learning_rate": 2.811131125388949e-05,
|
| 1829 |
+
"loss": 11.881485748291016,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.3062533530481661,
|
| 1834 |
+
"grad_norm": 24.61021614074707,
|
| 1835 |
+
"learning_rate": 2.7948128019857052e-05,
|
| 1836 |
+
"loss": 11.341839599609376,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.3112780183976027,
|
| 1841 |
+
"grad_norm": 20.106550216674805,
|
| 1842 |
+
"learning_rate": 2.7784817305509763e-05,
|
| 1843 |
+
"loss": 9.277299499511718,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.3163026837470395,
|
| 1848 |
+
"grad_norm": 27.051685333251953,
|
| 1849 |
+
"learning_rate": 2.762138617258346e-05,
|
| 1850 |
+
"loss": 10.385700988769532,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.3213273490964763,
|
| 1855 |
+
"grad_norm": 16.255699157714844,
|
| 1856 |
+
"learning_rate": 2.7457841688020996e-05,
|
| 1857 |
+
"loss": 10.452284240722657,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.3263520144459129,
|
| 1862 |
+
"grad_norm": 24.749526977539062,
|
| 1863 |
+
"learning_rate": 2.7294190923666708e-05,
|
| 1864 |
+
"loss": 9.209141540527344,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.3313766797953495,
|
| 1869 |
+
"grad_norm": 22.627124786376953,
|
| 1870 |
+
"learning_rate": 2.7130440955960557e-05,
|
| 1871 |
+
"loss": 8.883187866210937,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.3364013451447863,
|
| 1876 |
+
"grad_norm": 20.343547821044922,
|
| 1877 |
+
"learning_rate": 2.696659886563218e-05,
|
| 1878 |
+
"loss": 9.321096038818359,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.341426010494223,
|
| 1883 |
+
"grad_norm": 19.151782989501953,
|
| 1884 |
+
"learning_rate": 2.68026717373947e-05,
|
| 1885 |
+
"loss": 8.420301818847657,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.3464506758436596,
|
| 1890 |
+
"grad_norm": 14.526168823242188,
|
| 1891 |
+
"learning_rate": 2.6638666659638373e-05,
|
| 1892 |
+
"loss": 8.413037872314453,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.3514753411930962,
|
| 1897 |
+
"grad_norm": 22.384971618652344,
|
| 1898 |
+
"learning_rate": 2.6474590724124067e-05,
|
| 1899 |
+
"loss": 7.762445068359375,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.356500006542533,
|
| 1904 |
+
"grad_norm": 26.80672836303711,
|
| 1905 |
+
"learning_rate": 2.631045102567663e-05,
|
| 1906 |
+
"loss": 8.135292053222656,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.3615246718919698,
|
| 1911 |
+
"grad_norm": 15.913880348205566,
|
| 1912 |
+
"learning_rate": 2.614625466187809e-05,
|
| 1913 |
+
"loss": 9.075648498535156,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.3665493372414064,
|
| 1918 |
+
"grad_norm": 16.831798553466797,
|
| 1919 |
+
"learning_rate": 2.5982008732760737e-05,
|
| 1920 |
+
"loss": 9.587909698486328,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.371574002590843,
|
| 1925 |
+
"grad_norm": 15.664331436157227,
|
| 1926 |
+
"learning_rate": 2.5817720340500123e-05,
|
| 1927 |
+
"loss": 7.977745056152344,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.3765986679402797,
|
| 1932 |
+
"grad_norm": 17.52627182006836,
|
| 1933 |
+
"learning_rate": 2.5653396589107976e-05,
|
| 1934 |
+
"loss": 8.853733825683594,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.3816233332897165,
|
| 1939 |
+
"grad_norm": 16.778104782104492,
|
| 1940 |
+
"learning_rate": 2.548904458412495e-05,
|
| 1941 |
+
"loss": 9.27618408203125,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.3866479986391531,
|
| 1946 |
+
"grad_norm": 24.223648071289062,
|
| 1947 |
+
"learning_rate": 2.5324671432313452e-05,
|
| 1948 |
+
"loss": 8.344271087646485,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.3916726639885897,
|
| 1953 |
+
"grad_norm": 17.22234344482422,
|
| 1954 |
+
"learning_rate": 2.5160284241350278e-05,
|
| 1955 |
+
"loss": 7.873287200927734,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.3966973293380265,
|
| 1960 |
+
"grad_norm": 17.022199630737305,
|
| 1961 |
+
"learning_rate": 2.4995890119519315e-05,
|
| 1962 |
+
"loss": 7.8790740966796875,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.4017219946874633,
|
| 1967 |
+
"grad_norm": 17.03339385986328,
|
| 1968 |
+
"learning_rate": 2.4831496175404118e-05,
|
| 1969 |
+
"loss": 7.150788116455078,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.4067466600368999,
|
| 1974 |
+
"grad_norm": 18.490753173828125,
|
| 1975 |
+
"learning_rate": 2.466710951758059e-05,
|
| 1976 |
+
"loss": 7.228173828125,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.4117713253863367,
|
| 1981 |
+
"grad_norm": 20.83214569091797,
|
| 1982 |
+
"learning_rate": 2.450273725430955e-05,
|
| 1983 |
+
"loss": 7.958638000488281,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.4167959907357732,
|
| 1988 |
+
"grad_norm": 17.127973556518555,
|
| 1989 |
+
"learning_rate": 2.4338386493229375e-05,
|
| 1990 |
+
"loss": 7.392658996582031,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.42182065608521,
|
| 1995 |
+
"grad_norm": 15.541766166687012,
|
| 1996 |
+
"learning_rate": 2.4174064341048676e-05,
|
| 1997 |
+
"loss": 7.277698516845703,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.4268453214346466,
|
| 2002 |
+
"grad_norm": 16.299806594848633,
|
| 2003 |
+
"learning_rate": 2.4009777903238962e-05,
|
| 2004 |
+
"loss": 7.548030853271484,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.4318699867840834,
|
| 2009 |
+
"grad_norm": 12.783644676208496,
|
| 2010 |
+
"learning_rate": 2.3845534283727452e-05,
|
| 2011 |
+
"loss": 6.408290100097656,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.43689465213352,
|
| 2016 |
+
"grad_norm": 22.33506202697754,
|
| 2017 |
+
"learning_rate": 2.3681340584589823e-05,
|
| 2018 |
+
"loss": 6.378539276123047,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.4419193174829568,
|
| 2023 |
+
"grad_norm": 15.321292877197266,
|
| 2024 |
+
"learning_rate": 2.351720390574315e-05,
|
| 2025 |
+
"loss": 6.524259185791015,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
}
|
| 2028 |
+
],
|
| 2029 |
+
"logging_steps": 20,
|
| 2030 |
+
"max_steps": 9555,
|
| 2031 |
+
"num_input_tokens_seen": 0,
|
| 2032 |
+
"num_train_epochs": 3,
|
| 2033 |
+
"save_steps": 1000000000,
|
| 2034 |
+
"stateful_callbacks": {
|
| 2035 |
+
"TrainerControl": {
|
| 2036 |
+
"args": {
|
| 2037 |
+
"should_epoch_stop": false,
|
| 2038 |
+
"should_evaluate": false,
|
| 2039 |
+
"should_log": false,
|
| 2040 |
+
"should_save": true,
|
| 2041 |
+
"should_training_stop": false
|
| 2042 |
+
},
|
| 2043 |
+
"attributes": {}
|
| 2044 |
+
}
|
| 2045 |
+
},
|
| 2046 |
+
"total_flos": 1.318936666326912e+16,
|
| 2047 |
+
"train_batch_size": 1,
|
| 2048 |
+
"trial_name": null,
|
| 2049 |
+
"trial_params": null
|
| 2050 |
+
}
|